flytekitplugins.inference.NIM#
- class flytekitplugins.inference.NIM(secrets, image='nvcr.io/nim/meta/llama3-8b-instruct:1.0.0', health_endpoint='v1/health/ready', port=8000, cpu=1, gpu=1, mem='20Gi', shm_size='16Gi', env=None, hf_repo_ids=None, lora_adapter_mem=None)#
Methods
- Parameters:
- setup_nim_pod_template()#
Attributes
- base_url
- pod_template